Patients who meet DSM-III criteria for dysthymia are a prevalent, heterogeneous group whose illness frequently follows a chronic, disabling course. Controversy exists as to whether this is an affective disorder or a characterologic-neurotic illness. Our long-term objective is to demonstrate that some patients with dysthymia have an affective disorder, related to major depressive disorder, whose course is altered by long-term treatment with monoamine oxidase inhibitors. Our specific aim is to show that short-term benefit from MAOI in some dysthymic patients identifies a group of patients who will have long-term benefit from this medication. We also wish to determine the frequency of loss of antidepressant effect with long-term use of MAOI, and to explore methods of reversing this effect. Other areas to be examined include long-term course of side effects which emerge during short-term treatment such as sleeplessness, dizziness, sexual dysfunction, etc., and whether improvement in social and interpersonal functioning seen in short-term treatment persists during long-term treatment. We also wish to follow several biochemical parameters during long-term treatment. The proposed method includes randomly switching (double-blind) half of a group of dysthymic patients who have benefited from phenelzine for 6 months to placebo and maintaining half on phenelzine. Both groups will be evaluated for another 6 months. This will allow assessment of the long-term drug effect on mood and social functioning. Studies of the group maintained on MAOI for 12 months will allow assessment of the long-term course of side effects and the frequency of loss of antidepressant effect. The long-term health implications include demonstrating that there is a subgroup of patients with this common, chronic, pernicious illness whose illness course is altered with MAOI. Other data will shed light on the problems associated with long-term use of MAOI.